Introduction

These maps summarize the latest available COVID-19 outbreak data from the New York Time’s COVID-19 project, the City of St. Louis, and St. Louis County. The New York Times data are used for state and county-level data and local data sources are used for zip code and individual level data.

What Makes These Maps Unique?

Unlike other interactive maps being used to track the outbreak, the initial three maps include the counties in Illinois, Kansas, and Oklahoma that are part of Missouri’s metropolitan areas. Kansas City itself is also displayed here as if it were its own county. This is necessary because their public health department is reporting data for the city separate from the four counties that include parts of Kansas City.

The final map is also unique - it includes both the City of St. Louis and St. Louis County on one map and with a shared set of legend values, making direct comparisons possible. It shows Zip Code Tabulation Areas (ZCTAs), which are generalized areas that are roughly equivalent to USPS zip-codes. They are not 100% the same, however, and some homes with a given zip code may fall outside of ZCTA boundaries.

How to Use These Maps

These maps are fully interactive. Clicking on a county will reveal some details about that place. You can navigate around them just as you would with Google Maps. You can zoom with your mouse or the plus and minus buttons in the upper-left corner of each map. You may also move the maps around by clicking with your mouse and dragging.

Caveats

Due to lags in reporting, both at the public health department level and at the New York Times itself, these numbers may lag behind other figures reported in the media.

Prior to April 29th, I also used data from Johns Hopkins. As of the 29th, I am no longer using the Hopkins data because the New York Times only includes confirmed cases, while the Hopkins data inclues both confirmed and probable cases. Because probable cases are not being reported uniformly, and the New York Times is actively curating their data to remove them, their data seem to be the more stable (if conservative) option for the present.

Daily Data

While the City of St. Louis, St. Louis County, and Kansas City provide day-by-day tracking of cumulative cases on their respective dashboards, the State of Missouri does not. The following tabs provide daily breakdowns of COVIV data as well as additional statistics not included in the existing dashboards. Data are also provided for the wider St. Louis and Kansas City metro areas. Please note that the two average columns for reported cases and deaths are both seven-day rolling averages.

Missouri

St. Louis Metro

City of St. Louis

St. Louis County

Kansas City Metro

Kansas City

Daily Testing Data

Daily testing data are complied from a number of sources. For Missouri, the data set has been sourced directly from data posted by the Missouri Department of Health & Senior Services using a combination of data provided by Denis Beganovic, screen shots provided by The Alantic’s COVID Tracking Project, and the wayback machine. For other states, data is sourced from The Alantic’s COVID Tracking Project.

New Tests Average

Testing Rate

Percent Positive

Missouri Data

For Missouri, total numbers of test data are not available for many days in March and some days in April and May. For April and May, missing data are imputed using a linear teachnique. If Day 1 had 100 tests, and Day 3 had 200, Day 2 would assumed to have had 150 tests. Rates are per 100,000 residents, and averages are 7-day rolling averages.

All Data

For Missouri, total numbers of test data are not available for many days in March and some days in April and May. For April and May, missing data are imputed using a linear teachnique. If Day 1 had 100 tests, and Day 3 had 200, Day 2 would assumed to have had 150 tests. Rates are per 100,000 residents, and averages are 7-day rolling averages.

Health Care Infrastructure

This first map uses data from the Kaiser Health Network to identify counties (in gray) without any hospitals as well as the number of ICU beds per 1,000 residents in counties that do have hospitals. Keep in mind that some hospitals may have expanded ICU bed capacity in anticipation of increased need.

For Kansas City, all hospital and ICU bed data have been allocated to Jackson, Clay, Cass, and Platte Counties. If you have a sense of number of beds in Kansas City, MO itself as opposed to the surrounding counties, please feel free to drop me an email.


Infection Rates by County

This map shows infections as a rate per 1,000 residents. It is important not to map the raw counts themselves, but if you want to see those data, click on a county. You can also view the hospital infrastructure details from the first map for each county by clicking on them or by viewing the data table.

Map

Data Table

Average New Cases by County

This map shows a seven-day rolling average of new cases. For this map, this covers 2020-05-05 back through 2020-04-28. There is not a threshold for what constitutes a high or low average, but the higher the average number of new cases, the more new spread we can infer. For mapping purposes, these are displayed as a rate per 1,000 residents. As with the prior maps, additional details are available by clicking on each county or on the data table.

Map

Data Table

Mortality Rates by County

This map shows mortality as a rate per 1,000 residents. As with the prior maps, raw counts of deaths and hospital infrastructure details are available by clicking on individual counties or on the data table.

Map

Data Table

Infection Rates by St. Louis ZCTA (Zip Code)

This map shows infections as a rate per 1,000 residents for all ZCTAs with five or more patients. It is important not to map the raw counts themselves, but if you want to see those data, click on a ZCTA or the data table. If a ZCTA straddles the city and the county, and only has reported infection numbers in one entity, its estimated partial population will be displayed. Similarly, estimated partial populations for zip codes that straddle outlying counties are used.

Map

Data Table

Health Disparities in St. Louis

One striking facet of the COVID-19 outbreak is the effect it has had on communities that have been historically been marginalized in the United States, including African Americans, indigenous nations (the Navajo Nation has one of the highest infection rates in country), and people incarcerated in jails/prisons. What ties these groups together is not inherent biological risk. Rather, disinvestment at the community level means the agencies and organizations that serve these communities often lack the resources needed to prevent the spread of infectious disease. This is long-standing social and health disparities that potentially leave individuals more susceptible to severe illness. The following plots trace emerging health disparities in the St. Louis region both by poverty and race.

Poverty (Zip Code)

Race (Zip Code)

Race, Morbidity (Individual)

Race, Mortality (Individual)

Technical Notes

  • The Fisher breaks algorithm (Fisher 1958, Slocum et al. 2005) is used to identify legend categories
  • Estimated populations for the the four counties around Kansas City as well as Kansas City itself were created using areal weighted interpolation (see Prener and Revord 2019)
  • Estimated populations for partial zip codes are also created using areal weighted interpolation (see Prener and Revord 2019)